NeurIPS sold out before workshop results were even released. That is not a minor inconvenience. It is a structural failure in how one of the world's premier AI conferences allocates access, and it forces researchers, especially those presenting main-track papers, into a deeply unfair position. If you are one of those researchers, you are now making a calculus that has nothing to do with science: Is it worth staying for a conference you can only partially attend?
We think the answer is yes, but only if you are strategic about it. The main conference itself remains the densest networking opportunity of the year. The poster sessions, the hallway conversations, the spontaneous meetups over coffee, those are not workshop-exclusive. And for a researcher who has already secured a main-track paper, those interactions are where your work finds its audience. We have written before about how to write papers that connect with top conference audiences, and the truth is that the real connection happens after the paper is accepted: when someone reads your poster, asks a sharp question, and you find a collaborator or a job lead right there on the convention floor. That still happens without workshops.
What you lose by skipping the workshop days is real, but it is narrower than you might fear. Workshops are where niche communities form around specific subfields. They offer deep dives, not broad exposure. If your work sits at the intersection of two subcommunities, a workshop might be the only place that audience gathers. But for most researchers, the main conference days provide far more cross-disciplinary interaction. If you are trying to build a network that spans beyond your immediate area, the main conference is actually the better venue. We recently covered how researchers fund their first A* conference, and one consistent theme was that the best ROI comes from the days when everyone is in the same room, not when the crowd has fractured into twenty parallel sessions.
Our specific takeaway is this: Do not cut your stay short. Instead, treat the workshop days as optional bonus rounds. Use the main conference to establish five to ten meaningful conversations. Exchange contact information. Set up follow-up meetings during the workshop period with people you met earlier. That is a better use of your time than scrambling for a workshop registration that may never open up. And for the conference organizers: a system that does not guarantee main-track authors access to all components of the same event is a system that undervalues the researchers whose work fills the schedule. Fixing that is not a luxury. It is the bare minimum for a conference that wants to call itself inclusive.